J. Am. Chem. SOC.1995,117, 12364-12365
12364
Supramolecular Design by Covalent Capture. Design of a Peptide Cylinder via Hydrogen-BondPromoted Intermolecular Olefin Metathesis Thomas D. Clark and M. Reza Ghadiri* Departments of Chemistry and Molecular Biology The Scripps Research Institute, La Jolla, California 92037 Received August 9, 1995 Noncovalent chemical approaches to supramolecular design often exploit self-assembling and self-organizing molecular processes to establish equilibrium conditions for the production of thermodynamically stable aggregates of predetermined form and function.’ Such strategies are inherently efficient and have several distinguishing attributes, among which the most noteworthy are minimization of molecular information and synthetic effort through the use of modular subunits, high cooperativity in the assembly process, and built-in error-checking. However, the “equilibrium aggregates” produced in this way often possess low kinetic stabilities. Therefore, the utility of such processes could be greatly enhanced if desired noncovalent ensembles could be kinetically captured and stabilized by covalent-bondforming processes.2 Here we report one such approach in the context of the design and synthesis of a novel peptide-based cylindrical structure possessing a hydrophilic cavity and hydrophobic outside surface characteristics. Directed hydrogenbonding interactions in nonpolar organic solvents are used to effect structural self-assembly and self-organization of a flat ring-shaped peptide subunit into two thermodynamically favored interconverting antiparallel /3-sheet-like hydrogen-bonded cylindrical ensembles3 Hydrogen-bond-promoted 38-member ring forming olefin metathesis4 in the presence of (Cy3P)2C12Ru=CHCH=CPh2 is shown to be an effective means for the selective capture of the noncovalent peptide ensemble in the desired covalently stabilized form. The synthetic strategy employed in this study is based on the previously described design principles employed in the construction of peptide tubular ensembles3s5as well as recent advances in the design of remarkably chemoselective olefin (1) (a) Whitesides, G. M.; Simanek, E. E.; Mathias, J. P.; Seto, C. T.; Chin, D. N.; Mammen, M.; Gordon, D. M. Arc. Chem. Res. 1995,28, 3744. (b) Ghadiri, M. R. Adu. Mater. 1995, 7, 675-677. (2) For excellent reviews of templates in synthesis, see: (a) Anderson, S.; Anderson, H. L.; Sanders, J. K. M. Acc. Chem. Res. 1993, 26, 469475. (b) Lindsey, J. S.New J . Chem. 1991, 15, 153-180. For selected publications, see: (c) McCallien, D. W. J.; Sanders, J. K. M. J . Am. Chem. SOC. 1995, 117,6611-6612. (d) Amabilino, D. B.; Ashton, P. R.; Brown, C. L.; Cordova, E.; Godinez, L. A.; Goodnow, T. T.; Kaifer, A. E.; Newton, S. P.; Pietraszkiewicz, M.; Philp, D.; Raymo, F. M.; Reder, A. S.; Rutland, M. T.; Slawin, A. M. Z.; Spencer, N.; Stoddart, J. F.; Williams, D. J. J . Am. Chem. SOC. 1995, 117, 1271-1293. (e) Fujita, M.; Ibukuro, F.; Yamaguchi, K.; Ogura, K. J. Am. Chem. SOC. 1995, 117,4175-4176. (f) Chambron, J. C.; Dietrichbuchecker, C. 0.;Heitz, V.; Nierengarten, J. F.; Sauvage, J.-P.; Pascard, C.; Guilhem, J. Pure Appl. Chem. 1995,67, 233240. (g) Johnston, A. G.; Leigh, D. A.; Nezhat, L.; Smart, J. P.; Deegan, M. D. Angew. Chem., Int. Ed. Engl. 1995, 34, 1212-1216. (h) Wintner. E. A,; Conn, M. M.; Rebek, J. J . Am. Chem. SOC. 1994, 116, 8877-8884. (i) Li, T.; Nicolaou, K. C. Nature 1994,369,218-221. Q) Ghadiri, M. R.; Case, M. A. Angew. Chem., Int. Ed. Engl. 1993, 32, 1594-1597. (k) Harada, A.; Li, J.; Kamachi, M. Nature 1993, 364, 516-518. (1) Ghadiri, M. R.; Soares, C. J.; Choi, C. J . Am. Chem. SOC. 1992, 114, 4000-4002. (m) Terfort, A.; van Kiedrowski, G. Angew. Chem., Int. Ed. Engl. 1992, 31, 654-656. (n) Luebke, K. J.; Dervan, P. B. J . Am. Chem. SOC. 1989, 111, 8733-8735. (3)(a) Ghadiri, M. R.; Kobayashi, K.; Granja, J. R.; Chadha, R. K.; McRee, D. E. Angew. Chem., Int. Ed. Engl. 1995, 34, 93-95. (b) Kobayashi, K.; Granja, J. R.; Ghadiri, M. R. Ibid. 1995, 34, 95-98. (c) Sun, X.; Lorenzi, G. P. Helu. Chim. Acta 1994, 77, 1520-1526. (4) (a) Miller, S. J.; Grubbs, R. H. J . Am. Chem. SOC. 1995, 117,58555856. (b) Nguyen, S. T.; Grubbs, R. H. J . Am. Chem. SOC. 1993, 115, 9858-9859. (c) Fu, G. C.; Nguyen, S. T.; Grubbs, R. H. J . Am. Chem. SOC.1993, 115, 9856-9857. (d) Fu, G. C.; Grubbs, R. H. J . Am. Chem. SOC. 1992, 114, 5426.
metathesis catalysts4 (Figure 1). Cyclic peptides made up of an even number of alternating D- and L-amino acids bearing selective N-methylated peptide backbone amide functionalities have been shown to self-assemble in nonpolar organic solvents to form hydrogen-bonded cylindrical en~embles.~It was hypothesized that such ensembles, when endowed with intersubunit juxtapositioning of appropriate functionalities, should be able to readily participate in intermolecular covalent-bondforming processes to form stabilized supramolecular cylindrical structures. The eight-residue cyclic peptide cyclo[-(L-Phe-DMeNAla-~-Hag-~-MeNAla)2-] was designed for the task at hands6 The peptide ring structure was equipped with two L-homoallylglycine (Hag) residues for use in olefin metathesis reactions. The C2 symmetry of the peptide subunit (perpendicular to the plane of the ring structure) allows for the formation of two interconverting hydrogen-bonded diastereomeric ensembles 2a and 2b, which are distinguished by the identity of their crossstrand near-neighbor residues.3b It was hypothesized that the hydrogen-bonded ensemble 2a, which favors a productive orientation of the olefinic side chain moieties, would lower the activation entropy for the intermolecular carbon-carbon bond forming olefin metathesis process. In this way, the equilibrium would siphon through the kinetically labile ensemble 2a and convert into the covalently stabilized tricyclic structure 3. In nonpolar organic solvents, the peptide subunit selfassembles to form two slow-exchanging antiparallel #%sheetlike hydrogen-bonded cylinders with an association constant of Ka(CDC13) = 99 =k 1 M-’ (Figures 1 and 2a). Remarkably, in the presence of Grubbs ruthenium carbene the desired covalent supramolecular structure 3 is formed in 65% yield.8 The metathesis reaction gave three inseparable diastereomeric products, cic,cis, trans,trans, and cis,trans, as evidenced by ‘H NMR spectroscopy and ion-spray mass spectrometrys (Figure (5) (a) Granja, J. R.; Ghadiri, M. R. J . Am. Chem. SOC.1994,116,1078510786. (b) Ghadiri, M. R.; Granja, J. R.; Buehler, L. Nature 1994, 369, 301-304. (c) Khazanovich, N.; Granja, J. R.; McRee, D. E.; Milligan, R. A.; Ghadiri, M. R. J. Am. Chem. SOC. 1994,116,6011-6012. (d) Ghadiri, M. R.; Granja, J. R.; Milligan, R. A.; McRee, D. E.; Khazanovich, N. Nature 1993,366, 324-327. (6) Homoallylglycine (Hag) was prepared from ethyl acetamidocyanoacetate and 4-bromo-1-butene following the procedure of A l b e r t ~ o nand ~~ resolved (porcine kidney acylase I) following the procedures of Chenault et Standard Boc protection and purification by flash chromatograph gave the requisite amino acids having ’99% ee according to Marfey’s test.): The linear peptide (D-MeNAla-L-Hag-o-MeNAla-L-Phe)z was synthesized on commercially available N-Boc-L-Phe-PAM resin using standard Boc chemisyhcleavedfrom the solid support using tetrabutylammonium fluoride in DMF, and purified by RP-HPLC (44% yield). Linear peptide (1 mM) was cyclized at 0 “C in the presence of 3.0 equiv of l-hydroxy-7azaben~otriazole~~ (HOAt), 3.0 equiv of O-(7-azabenzo~azo1-1-y1)-1,1,3,3tetramethyluronium hexafl~orophosphate~~ (HATU) in 99.5% DMF, 0.5% v/v DIEA to give a 70% yield of the cyclic peptide following purification by RP-HPLC (IS-MS: MWaund = 857.4). (7)(a) Albertson, N. F. J . Am. Chem. SOC. 1946, 68, 450-453. (b) Chenault, H. K.; Dahmer, J.; Whitesides, G. M. J . Am. Chem. SOC. 1989, 111,6354-6364. (c) Marfey, P. Carlsberg Res. Commun. 1984.49.591596. (d) Ueki, M.; Kai, K.; Amemiya, M.; Horino, H.; Oyamada, H. J . Chem. Soc., Chem. Commun. 1988, 414-415. (e) Carpino, L. A. J . Am. Chem. SOC. 1993, 115, 4397-4398. (f) Carpino, L. A.; El-Faham, A.; Minor, C. A.; Albericio, F. J . Chem. Soc., Chem. Commun. 1994, 201203. (8) In a typical experiment, a dried solution of cyclic peptide (3-18 mg, 5.0 mM in CDCI,) and the ruthenium carbene catalyst (20-25 mol %) was allowed to react at room temperature under an Ar atmosphere for 48 h to give a 65% yield of the metathesis product 3 as a mixture of three olefin isomers, which were collectively isolated by RP-HPLC (MHlound= 1657.5). Hydrogenation in the presence of 1 wt equiv of 10% P d C in MeOH can be performed either on the purified products or directly on the crude reaction mixture to give quantitatively the reduced product 4 (MH” = 1661.7): ’H NMR (500 MHz, CDC13) 6 7.83 (d, J = 7.6 Hz, 4 H, NH), 7.56 (br, 4 H, Phe NH), 7.29-7.16 (m, 20 H, Phe), 5.16 (q, J = 7.0 Hz, 4 H, o-MeNAla(a)a-Ck), 5.12 (m, 4 H, Phe a-CH), 4.76 (m, 4 H, Hag a-CH), 4.59 (br, 4 H, o-MeNAla(b) a-CH), 3.06 (s, 12 H, D - ~ ~ NAla(b) NCH,), 3.04-2.92 (m, 8 H, Phe P-CHz), 2.48 (s, 12 H, D - ~ ~ NAla(a) NCHj), 1.54 (m, 8 H, H, Hag P-CHz), 1.40 (d, J = 7.3 Hz, o-MeNAla(b)B-CH3). 1.26-1.18 (m, 16 H, Hag, y f 6 CH2). 1.09 (d, J = 7.0 Hz,12 H, o-MeNAla(a)P-CH3).
wi
0002-786319511517-12364$09.00/0 0 1995 American Chemical Society
J. Am. Chem. SOC., Vol. 117, No. 49, 1995 12365
Communications to the Editor
b
7
-
9.0
1
,
I._
8.0
,
.
, 6.0
7.0
5.0
ppm
Figure 2. 'H NMR (500 MHz, CDCl3) spectra displaying NH and C, resonances of (a) peptide 1 (5.0 mM), (b) metathesis products 3 (7.1 mM at 293 K, mixture of three olefin isomers), and (c) product 4 (4.8 mM, at 283 K). Spectrum a displays non-hydrogen-bonded NH resonances for the peptide monomer 1 at BNH = 7.04 and 7.14 ppm
Self-Assembly
and hydrogen-bonded NH resonances for the two equally populated dimeric cylindrical ensembles 2a and 2b at BNH = 8.48,8.59, and 8.81 (2H). Spectrum b reflects the presence of the three diasteromeric (cic,cis, trans& and trans,trans) metathesis products. Reduction of the olefins produces a single symmetrical product with BNH = 7.56 and 7.83 ppm (spectrum c ) . ~
l
2a
t
3
1
HP,PdIC
9 Figure 1. Schematic representation of the synthetic strategy employed in the design of the tricyclic cylindrical peptide structure. The flat ring-shaped peptide subunit 1 self-assembles in nonpolar organic solvents to give an equal mixture of the hydrogen-bonded peptide ensembles 2a and 2b. Only ensemble 2a, which has the proper juxtaposition of olefinic side chains, can participate in the olefin metathesis which siphons the equilibrium toward the kinetically stable tricyclic structure 3. For clarity most side chains are omitted. D and L refer to the amino acid chirality.
2b). The reaction appears to proceed in a stepwise fashion as judged by the transient buildup of small quantities of an intermediate dimeric species where only the first intermolecular side chain-side chain olefin bond has formed. Purified samples of this intermediate, when subjected to the methathesis reaction conditions, convert to the tricyclic products 3. Reaction products 3 were hydrogenated in the presence of P d C in order to eliminate 'HNMR spectrum complexity due to the presence of olefin isomers. Reduction afforded, quantitatively, saturated
product 4, which was characterized by 'H NMR spectroscopy (ROESY, COSEY), mass spectrometry, and lT-IR spectroscopy.* The 'HNMR (CDCl3) spectrum is indicative of the expected symmetrical structure, having hydrogen-bonded N H resonances appearing at 7.83 and 7.56 ppm (Figure 2c). The I?"-IR spectrum displays amide-I resonances at 1632 and 1674 cm-I, closely resembling that of previously characterized peptide nanotube structure^?"^-^ and an NH stretching frequency at 3310 cm-I, which is also in agreement with the expected 4.7-4.8 8, hydrogen-bonded backbone-backbone di~tance.~ A requirement for intersubunithydrogen-bonding interaction for the promotion of olefin metathesis is supported by the following observations. When the reaction was conducted in polar organic solvents (1:1 DMF/CHC13 or EtOH), which are known from 'HNMR studies to disrupt intersubunit hydrogen-bonding interactions, no metathesis products were observed even after prolonged reaction times.'O Moreover, the directive effects of H-bonding interactions may have also contributed in preventing formation of undesired side products such as the intramolecularly metathesized peptide structure and oligomeric species as judged by ionspray mass spectrometry performed on the crude reaction mixture. In summary, we have presented a novel approach for the design and synthesis of a covalently stabilized B-sheet-like peptide cylindrical ensemble. The covalent capture strategy presented here is expected to be useful in stabilizing kinetically labile a-helical and /$sheet peptide secondary structures and may also find utility in de novo design of artificial proteins. Acknowledgment. We thank colleagues D. Hwang and M. Case for assistance in NMR analysis, A. Hoveyda and M. Visser for providing a sample of the ruthenium carbene catalyst, and R. Grubbs and S . Miller for communicating their results prior to publication." Financial support for this work was provided by the National Institutes of Health (GM 52190). T.D.C. is an NSF Predoctoral Fellow and M.R.G. an Alfred P. Sloan Research Fellow 1993-1995. JA9527176 ~~
~~
~
~
~~~
(9) Krimm. S.; Bandekar. J. hAduances in Protein Chemisfn;Anfinsen, C. B.. Edsall, J. T.; Richards, F. M., Eds.; Academic Press: Orlando, 1986; pp 181-364. (IO)It should be noted that these solvent mixtures can also drastically reduce the activity of the catalyst (Grubbs, R.: Miller, S. Personal communication).